What is security awareness?
Definition: Security awareness is the knowledge, attitude and habitual behaviour that lead employees to recognise cybersecurity threats, follow security policies and respond appropriately to suspicious activity, for example by verifying an unusual request or reporting a suspicious message.
Security awareness has three layers. Knowledge is knowing that an attack type exists and what it looks like. Attitude is caring enough to act on that knowledge when busy. Behaviour is what people actually do when a convincing email, call or message arrives. Programmes aim at the third layer: a person who passes a quiz but still approves a fake invoice has knowledge without the behaviour.
Awareness is therefore different from completing a course. A completed module shows that someone was exposed to information; it does not show that they will act differently. The US standards body NIST takes the same view. Its guidance SP 800-50 Rev. 1 (2024) treats awareness activities, practical exercises and training as parts of one learning programme that should encourage behaviour change as part of risk management.
Security awareness is a continuous process. Attack methods change, people change roles, new staff join and habits fade without practice. For that reason, most programmes combine a yearly baseline with short, regular activities throughout the year.
Security awareness and related terms
| Term | What it means | How it relates to security awareness |
|---|---|---|
| Security awareness | The state of recognising threats and acting securely | The goal |
| Security awareness training | Structured teaching of threats, policies and correct responses | One means of building awareness |
| Cybersecurity awareness | Often used as a synonym; sometimes broader, including private online safety | Largely interchangeable in a workplace context |
| Phishing simulation | A realistic, harmless test attack with immediate feedback | Practice and measurement of behaviour |
| Human risk management | Using behavioural and risk data to prioritise interventions | Extends awareness from teaching to risk-based steering |
| Security culture | Shared norms and beliefs about security in an organisation | The environment that makes secure behaviour normal |
Why is security awareness important?
Security awareness is important because many cyberattacks succeed by manipulating people rather than breaking technology. Verizon's 2025 Data Breach Investigations Report found human involvement in 60% of the breaches it analysed, and ENISA's Threat Landscape 2025 names phishing, including vishing, as the leading way into organisations, at 60% of the EU incidents it analysed between July 2024 and June 2025. Technical controls such as email filtering and multi-factor authentication reduce these attacks but do not stop all of them, so the person who receives the message often decides whether an attack succeeds or is reported early.
Framing employees as the problem is neither accurate nor useful. People are also the sensors that notice an unusual request, a strange login prompt or a caller who knows too much. A good programme strengthens that detection role.
Threats that depend on human behaviour
- Social engineering: manipulating people into revealing information or taking actions, using authority, urgency, helpfulness or fear.
- Phishing and credential theft: messages that lead to fake login pages or malicious attachments. Some modern techniques, such as adversary-in-the-middle proxies, can bypass weaker forms of multi-factor authentication.
- Business email compromise (BEC) and CEO fraud: impersonating executives, suppliers or partners to redirect payments. The FBI's Internet Crime Complaint Center recorded 24,768 BEC complaints with reported losses of about 3.05 billion US dollars in 2025 (IC3 2025 report). Our analysis of CEO fraud losses in Germany covers the German figures.
- Voice phishing (vishing): phone calls that impersonate IT support, banks or managers, often to obtain passwords or approve access.
- Deepfake impersonation: synthetic voice or video of a real person. In 2024, an employee of the engineering firm Arup in Hong Kong transferred about 25 million US dollars after a video call in which the other participants were deepfakes (CNN).
- Accidental data exposure: sending data to the wrong recipient, sharing files publicly or pasting confidential information into unapproved tools.
- Unsafe handling of sensitive information: weak passwords, unlocked devices, discussing confidential matters in public or using personal accounts for work data.
- Late or missing incident reporting: an attack that is noticed but not reported has more time to spread.
Malicious insiders and unintentional error
Security awareness mainly addresses unintentional behaviour: well-meaning people who are deceived, rushed or unsure of the correct process. Deliberate malicious insiders are a different risk. They are addressed mainly through access control, monitoring, segregation of duties and HR processes, although an aware workforce is more likely to notice and report suspicious behaviour by colleagues.
What does security awareness include?
A modern security awareness programme includes four connected activities: educating employees, letting them practise with realistic simulations, making it easy to report suspicious activity, and measuring results to improve the programme. We describe these as Educate → Simulate → Report → Measure and improve. This is an explanatory framework used in this guide, not a formal standard; regulations and standards describe the same elements in their own terms.
| Component | Purpose | Typical activities |
|---|---|---|
| Educate | Build knowledge of threats, policies and correct responses | E-learning, microlearning, workshops, role-specific modules |
| Simulate | Practise recognising attacks under realistic conditions | Phishing, vishing, smishing and deepfake simulations with feedback |
| Report | Turn recognition into action that protects others | Report buttons, phone or messaging reporting, clear escalation paths |
| Measure and improve | Show what changes and target the next activities | KPIs, risk trends by group, programme reviews |
Security awareness training is the structured teaching element. E-learning is its most common delivery format. Phishing and social engineering simulations test and practise behaviour. Behavioural feedback is the short explanation a person receives right after a simulation, while the situation is still fresh. Threat reporting is the action of notifying the security team about a suspicious message, call or event. Human risk management uses data from all of these to decide where to focus, and continuous improvement closes the loop.
Security awareness training methods
No single training method is best for every goal. Short, frequent formats suit habit-building, interactive scenarios suit complex decisions, and live sessions suit leadership and high-risk teams. Most programmes combine several methods.
| Method | Strengths | Limitations | Suitable for |
|---|---|---|---|
| Traditional e-learning | Scalable, consistent, easy to document | Passive; easily clicked through; quickly outdated | Baseline and compliance topics |
| Microlearning | Short (often under five minutes); fits work schedules | Limited depth per unit | Regular reinforcement of single behaviours |
| Interactive training | Requires decisions; higher engagement | More effort to produce | Practising judgement |
| Scenario-based learning | Realistic context; shows consequences | Needs maintenance to stay current | Fraud processes, payment approvals, incident response |
| Gamification | Motivation through points, levels or team goals | Can reward speed over care if badly designed | Sustaining participation over time |
| Role-specific training | Matches real exposure and authority | Requires role mapping | Finance, HR, IT helpdesk, executives, contact centres |
| In-the-moment learning | Feedback when a mistake or near-miss happens | Can feel intrusive if poorly framed | Follow-up to simulations and real events |
| AI-personalised learning | Adapts difficulty and topics to the individual or role | Depends on data quality and privacy safeguards | Large, diverse workforces |
| Workshops and live training | Discussion, questions, trust | Costly; hard to scale | Leadership, high-risk teams, onboarding |
Design choices matter more than labels. Our article on which gamification mechanics change behaviour shows how the same format can help or backfire depending on what it rewards.
Phishing and social engineering simulations
A phishing simulation is a realistic but harmless attack sent to employees with the organisation's authorisation, followed by immediate feedback. It lets people practise recognising and reporting attacks in their normal working environment, and it gives the organisation data on how behaviour develops. Our phishing simulation guide explains planning and operation in detail.
A simulation is different from an educational demonstration. A demonstration shows employees what an attack looks like, for example a deepfake video in a training module; a simulation tests whether they recognise one when it arrives unannounced.
Types of simulation
| Simulation type | What it imitates | What it practises |
|---|---|---|
| Email phishing | Malicious links, attachments, fake login pages | Checking senders, links and requests |
| Spear phishing | Messages tailored to a person or role, often using OSINT | Recognising personalised pretexts |
| Vishing | Phone calls from fake IT, banks or managers | Verifying callers; refusing to share credentials |
| Smishing | SMS with links or callback numbers | Treating texts with the same caution as email |
| Messaging and collaboration attacks | Messages in WhatsApp, Teams or similar | Verifying identities outside email |
| Quishing | QR codes leading to phishing sites | Checking where a QR code leads |
| Deepfake voice | A cloned voice of a known person | Verifying unusual requests through a second channel |
| Deepfake video | A synthetic person in a video call | Following verification rules even when a face looks familiar |
| Multi-channel scenarios | An email followed by a call or message | Recognising coordinated pretexts |
Our articles on vishing and AI voice phishing, how voice and video deepfakes are made and deepfake simulations under GDPR and the EU AI Act cover these channels in depth.
How simulations work
- Plan: define objectives, target groups, channels and frequency, and agree rules with data protection and, where one exists, the works council.
- Design: build scenarios that reflect real threats to the organisation and match each group's role and experience.
- Deliver: send simulations over time rather than as a single campaign, so results reflect normal working conditions.
- Give feedback: show the person what the clues were and what to do next time, immediately and without blame.
- Measure: record interactions, credential submissions and reports, ideally aggregated by group.
- Adapt: adjust difficulty, topics and channels to results and to new attack methods.
Simulations measure behaviour in a test situation. They show who interacts with a lure, who submits data and who reports it, and how quickly. They do not measure how people would react to every real attack, so results are best read as trends.
Ethics, privacy and employee trust
Simulations involve personal data and can damage trust if handled badly. Good practice includes informing employees that simulations take place, avoiding scenarios that exploit personal distress (for example fake bonus cuts or health scares), not naming or shaming individuals, keeping results aggregated where possible, and limiting retention. Some public guidance is sceptical of simulations altogether: the UK National Cyber Security Centre warns that they can erode trust between employees and security teams and that no training can teach users to spot every phishing attempt. Simulations therefore work best as one layer alongside technical controls and easy reporting, not as the main defence.
In the EU, the GDPR applies, and in Germany co-determination rights of the works council are likely to be relevant. Our GDPR checklist for phishing simulations and works agreement template cover the details. This is not legal advice.
Human risk management and security behaviour
Definition: Human risk management (HRM) is the practice of identifying, measuring and reducing cybersecurity risk that arises from human behaviour, by combining behavioural data, role-based exposure and targeted interventions.
Traditional security awareness asks "Has everyone been trained?". Human risk management asks "Where is human-related risk highest, and which intervention reduces it?". It builds on awareness data but adds prioritisation. A payment clerk who approves transfers, an IT helpdesk agent who resets passwords and an executive whose voice is public carry different exposure, even with the same training record.
Human risk management typically covers:
- Risk identification: which roles have authority over payments, credentials or sensitive data, and which are publicly visible.
- Risk assessment: combining exposure with behavioural indicators to set priorities.
- Behavioural indicators: simulation interactions, reporting behaviour, training engagement and, where appropriate, security events.
- Positive security behaviour: verifying unusual requests, reporting suspicious activity, using approved tools.
- Threat reporting: a measurable behaviour that also provides early warning to the security team.
- Improvement over time and security culture: whether secure behaviour becomes the norm across groups.
Human risk management should not become surveillance or a way of labelling individuals as risky. Aggregated, group-level analysis with clear purpose limits supports both privacy and trust. In practice, a high-risk role is a reason for better support, not for suspicion of the person.
Measuring security awareness effectiveness
Security awareness is effective when behaviour changes, not when courses are completed. Training completion shows reach and is useful for compliance records, but it says nothing about whether people recognise and report attacks. Behavioural metrics from simulations and real reports give a fuller picture, especially when tracked as trends over time.
| KPI | What it measures | Calculation | What it can indicate | Limitations |
|---|---|---|---|---|
| Training completion rate | Reach of training | Completed ÷ assigned | Coverage; compliance evidence | Says nothing about behaviour |
| Phishing interaction rate | Engagement with simulated lures | Interactions (clicks, replies, QR scans) ÷ delivered | Susceptibility to a scenario | Strongly depends on scenario difficulty |
| Credential submission rate | Most serious simulated failure | Submissions ÷ delivered | Risk of account takeover | Low numbers; volatile |
| Simulation reporting rate | Protective action | Reports ÷ delivered | Detection capability | Can be inflated by reporting everything |
| Time-to-report | Speed of protective action | Time from delivery to first report (or median of all reports) | How fast the security team could be warned | Define which variant you use |
| Correct reporting | Accuracy of reports | Reports of genuine threats ÷ all reports | Quality of judgement | Needs triage data |
| Reporting across channels | Breadth of protective behaviour | Reports by channel (email, call, SMS, messaging) | Whether awareness extends beyond email | Requires reporting options in each channel |
| Behavioural change over time | Programme effect | Trend of the metrics above per period | Whether the programme works | Needs consistent scenarios to compare |
| Risk trends by group | Where to focus | Metrics aggregated by department, role or site | Targeting of interventions | Small groups can identify individuals; set a minimum group size |
Simulated outcomes are not the same as real incidents. A falling interaction rate in simulations is encouraging, but the stronger evidence is a rising number of timely, accurate reports of real threats and fewer incidents with a human component.
There is no universal benchmark for a "good" click rate, because results depend heavily on scenario difficulty, targeting and channel. Peer-reviewed field studies also show that effects vary. A study of more than 14,000 employees over 15 months found that training embedded in simulations did not make employees more resilient, while a report button made fast, crowd-sourced detection of new campaigns practical (Lain et al., IEEE S&P 2022). A randomised study of more than 19,500 employees at UC San Diego Health found no significant link between recent annual training and failing simulations, and only a 2% reduction in failure rates from embedded training (Ho et al., IEEE S&P 2025). On the reporting side, Verizon's 2025 DBIR found that recently trained users reported simulated phishing at about 21%, against a base rate of 5%. These findings argue for measuring reporting and real-world outcomes rather than relying on one metric. Our pilot scope template and success KPIs and the guide to evaluating social engineering training show how to set up measurement in practice.
Security awareness and compliance
Several regulations and standards require security awareness training and, in some cases, evidence that measures are effective. Training and simulation platforms support these activities and produce documentation, but they cannot guarantee compliance on their own: scope, governance and documentation remain the organisation's responsibility. This section is a summary, not legal advice.
| Framework | Relevant requirement | Implication for awareness programmes |
|---|---|---|
| NIS2 Directive | Art. 21(2)(g): basic cyber hygiene practices and cybersecurity training; Art. 21(2)(f): procedures to assess the effectiveness of risk-management measures; Art. 20(2): management bodies must follow training | Training for all staff and management; evidence of effectiveness |
| German BSI Act (NIS2 implementation) | § 30(2) No. 6 and 7 BSIG: effectiveness assessment, basic training and awareness measures; § 38(3) BSIG: regular training for management | Applies to entities in scope in Germany |
| DORA | Art. 13(6): ICT security awareness programmes and resilience training as compulsory modules for all staff, including senior management; Art. 5(4): specific training for the management body | Applies to financial entities; role-based content and records |
| ISO/IEC 27001:2022 | Clause 7.3: awareness; Annex A control 6.3: information security awareness, education and training; clause 9.1: monitoring, measurement, analysis and evaluation | Documented awareness and evidence of monitoring |
| GDPR | Art. 39(1)(b): the data protection officer monitors compliance, including awareness-raising and training of staff involved in processing; Art. 32: appropriate technical and organisational security measures | Training staff who process personal data; data protection rules for simulations themselves |
| BSI IT-Grundschutz | Module ORP.3 "Sensibilisierung und Schulung zur Informationssicherheit" (awareness and training) in the IT-Grundschutz-Kompendium | German reference for structuring awareness and training |
Two points are easy to miss. First, regulations require training and evidence; phishing simulations are a common method of producing that evidence, but none of the frameworks above prescribes them. Second, the awareness programme itself processes employee data and must comply with data protection law. For details, see our guides to NIS2 awareness training requirements in Germany, NIS2, ISO 27001 and BSI IT-Grundschutz and DORA and phishing simulations.
How to build an effective security awareness programme
An effective security awareness programme starts from the organisation's actual risks, sets behavioural goals, and improves through measurement. The eight steps below form a practical sequence; in a running programme, steps 4 to 8 repeat continuously.
| Step | What to do | Output |
|---|---|---|
| 1. Assess risks and current awareness | Review incidents, threat intelligence, critical processes and a baseline simulation | Risk picture and baseline |
| 2. Define objectives | Set behavioural goals, such as "verify every change of bank details by phone" | Measurable targets |
| 3. Segment employees | Group by role, authority and exposure | Target groups |
| 4. Select training methods | Match formats to groups and goals | Training plan |
| 5. Introduce realistic simulations | Start simple, add channels and difficulty over time | Simulation calendar |
| 6. Establish simple reporting | One-click report button, clear phone and messaging route, fast feedback | Working reporting process |
| 7. Measure outcomes | Track the KPIs above, by group and over time | Dashboard and reports |
| 8. Improve continuously | Update content for new threats; target groups that need support | Revised plan |
Assess and define. Start with the processes attackers target: payments, supplier data, password resets, access approvals. A behavioural objective is specific and observable; "be more careful" is not.
Segment and select. Finance, HR, IT helpdesk, contact centres, executives and their assistants usually need additional, role-specific scenarios. Frontline staff without a desk need formats that work on mobile devices or in short briefings.
Simulate and report. Introduce simulations gradually and announce the programme in general terms. Make reporting easier than ignoring: if reporting a suspicious call requires a ticket form, few people will do it.
Measure and improve. Review results quarterly, compare trends rather than single campaigns, and retire content that no longer reflects real attacks.
Prerequisites. Visible management support signals that security is part of the job. Early involvement of the data protection officer and, in Germany, the works council avoids delays and builds trust. Clear communication explains why the programme exists, what data is collected and that the aim is learning, not punishment.
AI and the future of security awareness
Artificial intelligence lowers the cost of convincing social engineering and changes what employees need to recognise. Fluent, error-free messages, cloned voices and synthetic video mean that spelling mistakes and unfamiliar voices are no longer reliable warning signs. Verification processes become more important than spotting flaws.
Established today:
- AI-generated phishing: attackers use language models to write fluent, localised messages at scale. In a 2024 study with 101 participants, fully AI-automated spear phishing emails achieved a 54% click-through rate, the same as emails written by human experts and well above the 12% of a generic control group (Heiding et al., 2024).
- Deepfake voice and video impersonation: voice cloning from short recordings and synthetic video calls are documented in real fraud cases, as the Arup case shows.
- OSINT-driven and hyper-personalised attacks: public profiles, company websites and leaked data supply the details that make a pretext believable. See what OSINT-based social engineering training is.
Established on the defence side:
- AI-assisted simulation creation: generating realistic, current scenarios and localising them quickly.
- Adaptive learning: adjusting difficulty and topics to each person or group.
- Threat-intelligence-driven training: turning current attack patterns into simulations and lessons within a short time.
Emerging:
- Social engineering of AI agents: as organisations deploy AI assistants with access to email, files or systems, attackers try to manipulate them through prompt injection. Employees need to understand which instructions their AI tools may act on.
- AI literacy requirements: Article 4 of the EU AI Act, applicable since 2 February 2025, requires providers and deployers of AI systems to take measures on the AI literacy of their staff. The article was amended in 2026, so check the current wording when planning training.
Claims that AI will make training obsolete, or that every attack will soon be a deepfake, are speculative. The more defensible expectation is that verification habits, fast reporting and process controls (such as call-back rules for payment changes) will matter more as fakes become harder to spot. Our article on why AI phishing breaks traditional training discusses the implications.
Common security awareness mistakes
Treating training as a compliance checkbox. A yearly course satisfies an auditor's question but rarely changes behaviour. Pair the baseline with regular practice and behavioural measures.
Measuring only completion rates. Completion shows reach, not effect. Add reporting rate, time-to-report and trends by group.
Relying only on generic email phishing. Attacks also arrive by phone, SMS, messaging apps and video. Include the channels your organisation actually faces, and offer a way to report in each.
Punishing or humiliating employees. Public shaming, disciplinary threats for clicking or deceptive scenarios built on personal distress damage trust and discourage reporting. The UK National Cyber Security Centre explicitly advises against a punishment or blame-oriented culture around phishing.
Ignoring positive reporting behaviour. If reporting receives no feedback, people stop doing it. Acknowledge reports and tell people what happened.
Giving everyone the same training. A payment approver and a warehouse worker face different attacks. Role-specific content is more relevant and more efficient.
Failing to update content. Scenarios based on attacks from several years ago teach outdated signals. Review content against current threat intelligence.
Neglecting culture and workflow. If a process makes verification difficult, for example no time to call back before a payment deadline, training cannot compensate. Fix the process and train the behaviour together.
Conclusion
Security awareness is a continuous discipline, not a course. It works when people know the threats that matter to their role, practise recognising them in realistic conditions, can report suspicious activity in a few seconds, and see that their reports make a difference. Measuring behaviour and reporting over time, rather than completions, shows whether a programme reduces risk and gives regulators the evidence they ask for.
For definitions of individual terms, see the revel8 glossary; for practical planning, our phishing simulation guide is a good next step.




